The choice of a Database Management System (DBMS) is a complex process because of their diversity and the various functionalities and performances they offers. The amount of available DBMSs makes it harder for decision-makers and database developers to find the most suitable DBMS according to their business requirements.
The aim of this thesis is to help them to make this choice. To do so, a detailed approach is proposed to make the selection process easier. This approach presents the advantages and weaknesses in terms of flaws and potential risks of the chosen system.
This approach is divided into three steps. First, we focus on deployment requirement identification on a database. Then, we evaluate the DBMS according to different criteria, related to the database nature, including the data volume, type and the query load. The goal of this evaluation is to analyze and study eligible DBMS features. These two steps will lead to the final one which is the elaboration of a comparison matrix to facilitate the choice of DBMS.
Keywords: Database Management Systems (DBMS), relational Database, classification, NoSQL, benchmarking.